Ruddiman, W., Sarnthein M., et al., 1989 Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 108 8. COMPARISON OF UPPER PLIOCENE DISCOASTER ABUNDANCE VARIATIONS FROM NORTH ATLANTIC SITES 552, 607, 658, 659, AND 662: FURTHER EVIDENCE FOR MARINE PLANKTON RESPONDING TO ORBITAL FORCING1 Alex Chepstow-Lusty,2 Jan Backman,3 and Nicholas J. Shackleton2 ABSTRACT Abundance variations of six Pliocene species of discoasters have been analyzed over the time interval from 1.89 to 2.95 Ma at five contrasting sites in the North Atlantic: Deep Sea Drilling Project Sites 552 (56°N) and 607 (41°N) and Ocean Drilling Program 658 (20°N), 659 (18°N), and 662 (1°S). A sampling interval equivalent to approximately 3 k.y. was used. Total Discoaster abundance showed a reduction with increasing latitude and from the effects of upwelling. This phenomenon is most obvious in Discoaster brouweri, the only species that survived over the entire time interval studied. Prior to 2.38 Ma, Discoaster pentaradiatus and Discoaster surculus are important components of the Discoaster assemblage: Discoaster pentaradiatus increases slightly with latitude up to 41°N, and its abundance relative to D. brouweri increases up to 56°N; D. surculus increases in abundance with latitude and with upwelling conditions relative to both D. brouweri and D. pentaradiatus and is dominant to the latter species at upwelling Site 658 and at the highest latitude sites. Discoaster asymmetricus and Discoaster tamalis appear to increase in abundance with latitude relative to D. brouweri. Many of the abundance changes observed appear to be connected with the initiation of glaciation in the North Atlantic at 2.4 Ma. The long-term trend of decreasing Discoaster abundance probably reflects the fall of sea-surface temperatures. This trend of cooling is overprinted by short-term variations that are probably associated with orbital forcing. Evidence for the astronomical elements of eccentricity and obliquity periodicities were found at all sites; however, only at Sites 607, 659, and 662 were precessional periodicities detected. Furthermore, only at Site 659 was precession found to be dominant to obliquity. Abundance peaks of individual species were found to cross-correlate between sites. The distinct abundance fluctuations observed especially in the tropics suggest that temperature is not the only factor responsible for this variation. This study reveals for the first time the importance of productivity pressure on the suppression of Discoaster abundance. INTRODUCTION In the present study, individual species and total abun• dances of discoasters are compared among five sites across a Discoasters are known in the oceanic record from the wide latitudinal range of the North Atlantic. We hope to upper Paleocene to the upper Pliocene, when the last two extend this research approach to other oceans and develop a representatives of the genus disappeared synchronously just global data base. As more sites are analyzed and compared, below the base of the Olduvai Subchron, 1.89 Ma ago Discoaster abundance variations may give us a global record (Takayama, 1970; Rio, 1982; Backman and Shackleton, 1983). of ocean temperature changes and/or nutrient levels. Numerous studies have shown that discoasters are particu• In addition, after we obtain continuous quantitative estimates larly common in low-latitude sediments and therefore proba• of Discoaster abundances, we will also be able to use spectral bly favored warm-water masses (e.g., Haq and Lohmann, analysis to determine whether orbital forcing is imprinted on the 1976; Bukry, 1978). The organism from which discoasters Discoaster record at these sites. The time interval investigated is evolved is not known, but it was presumably related to the approximately the million years preceding the extinction of the Coccolithophorids. Discoasters are favorable for quantitative discoasters (i.e., between 1.9 and 2.9 Ma) using closely spaced study since they resist dissolution better than planktonic samples (around 3 k.y.). The taxonomy of the six species of foraminifers or most placoliths (Lohmann and Carlson, 1981). discoasters counted (Plate 1) and the methods used in this study A number of factors probably affected their distribution, follow those of Backman and Shackleton (1983). but this genus appears especially sensitive to temperature change (Haq and Lohmann, 1976). Backman et al. (1986) and SITES SELECTED Backman and Pestiaux (1987) have shown that orbital forcing The five sites chosen (Fig. 1 and Table 1) were picked to is strongly imprinted on Discoaster abundance variation at contrast with each other either in terms of latitude, upwelling Deep Sea Drilling Project (DSDP) Hole 552A (56°N) and DSDP vs. nonupwelling systems, and the effects of cool currents. Site 606 (37°N) during the Pliocene in the North Atlantic. The However, all sites are from the eastern North Atlantic, which former is dominated by the 41-k.y. obliquity periodicity and the may limit our interpretations. The sites possess good se• latter by the 400-k.y. eccentricity periodicity. quences of upper Pliocene sediment and, except for Hole 552A, fairly high sedimentation rates. Sampling was at 10-cm intervals except for Sites 607 and 658, which were at 15-cm 1 Ruddiman, W., Sarnthein, M., et al., 1989. Proc. ODP, Sci. Results, 108: and 30-cm intervals, respectively. College Station, TX (Ocean Drilling Program). Godwin Laboratory, Subdepartment of Quaternary Research, University Site 659 of Cambridge, Cambridge CB2 3RS, United Kingdom. 3 Department of Geology, University of Stockholm, S-10691 Stockholm, This site was selected as the nonupwelling, open-ocean Sweden. reference site. It is located on the top of the Cape Verde 121 A. CHEPSTOW-LUSTY, J. BACKMAN, N. J. SHACKLETON 60 Table 1. North Atlantic sites selected, their geographic positions, and water depth. ROCKALL Depth Hole Location (m) Reference 552 662A PS, 11°W 3,824 Ruddiman, Sarnthein, et al., 1988 659A 18°N, 21°W 3,067 Ruddiman, Sarnthein, et al., 1988 658A 20°N, 18°W 2,263 Ruddiman, Sarnthein, et al., 1988 607 41°N, 33°W 3,427 Ruddiman, Kidd, Thomas, et al., 1987 552A 56°N, 23°W 2,304 Roberts, Schnitker, et al., 1984 50- NORTH ATLANTIC OCEAN production and eolian input. Discoasters may have fluctuated in response to the intensity of upwelling. Site 662 This site is located close to the equator on the upper eastern flank of the mid-Atlantic Ridge. The relevant interval is a nannofossil and foraminifer-nannofossil ooze. One would expect discoasters to be extremely abundant in such low 607 latitudes, although this site is influenced by the South Equa• 40- torial Current, which is the successor to the Benguela Cur• rent. It is within the equatorial divergence and may be affected by upwelling. AZORES Site 607 This site is presently located along the northernmost part of the subtropical gyre; however, in the late Pliocene, it was 30°- CANARY ISLANDS exposed to major circulation changes and lower temperatures (Ruddiman, Sarnthein, et al., 1988). The lithology is interbed• ded foraminifer-nannofossil oozes and marls representing the North Atlantic glacial marine cycles. Site 552 This site is at the northernmost latitude of our transect. The 20- data for this site was published by Backman and Pestiaux (1987) and has been replotted to provide a comparison with the other sites. It was drilled on the southwest margin of the Rockall Plateau and the lithology is similar to Site 607 in that CAPE VERDE it consists of cycles of alternating beds of foraminifer-nanno• ISLANDS fossil ooze and calcareous marls and mud that began abruptly 10- about 2.4 Ma. Prior to this age, the lithology is relatively uniform foraminifer-nannofossil ooze and nannofossil ooze. AGE MODELS AND SEDIMENTATION RATES In the absence of magnetostratigraphic data from Sites 659 and 662 and imprecise data from Site 658, all age models are 0- based on biostratigraphic datums. However, at Sites 552 and SOUTH ATLANTIC OCEAN 662 607 very reliable magnetic stratigraphies were obtained. Only three stratigraphic datums have been selected as control 30° W 20° 10* points for the age models at each site. Where it was possible, Figure 1. Location of sites studied in the North Atlantic. biostratigraphic datums were selected preferentially, espe• cially those related to Discoaster events (Table 2). This ensures that the age models from the five sites are self- Plateau. The oceanographic setting is not complicated by the consistent. cool Canaries Current or by large inputs of terrigenous matter At Site 659 magnetic data was available for most of Hole from northwest Africa. The lithology is mostly nannofossil- 659B, but magnetic reversals were unambiguous only in the foraminifer ooze with good preservation and no noticeable upper section of Hole 659A. Using distinct D. brouweri dissolution in the interval studied. abundance peaks for correlation, it was possible to switch from Hole 659A to Hole 659B to obtain the extinction event, Site 658 which was not recovered in Hole 659A. Most of the species In contrast, Site 658 was the upwelling, high-productivity events fit a single line on the age-depth plot (Fig. 2). To site chosen in close proximity to Site 659. It is located on the develop the age model, we used the datums of the extinction continental slope 160 km west of Cap Blanc and lies beneath of D. brouweri and D. triradiatus at 1.89 Ma, the base of the a major productive cell of oceanic upwelling. The sediments D. triradiatus acme at 2.07 Ma (Backman and Shackleton, are hemipelagic due to the combined influence of oceanic 1983), and the extinction of D. tamalis at 2.65 Ma (Table 3). 122 COMPARISON OF UPPER PLIOCENE DISCOASTER ABUNDANCE VARIATIONS Table 2. Summary of Discoaster events between 1.89 Table 3. Summary of control points used in age models. and 3.00 Ma from previous work (e.g., Backman and Shackleton, 1983; Backman et al., 1986).